/ . Embryol. exp. Morph. Vol. 40, pp. 199-214, 1977
Printed in Great Britain © Company of Biologists Limited 1977
J99
Somatic nuclei in amphibian oocytes: evidence
for selective gene expression
By E. M. DE ROBERTIS, 1 G. A. PARTINGTON, 1
R. F. LONGTHORNE, 1 AND J. B. GURDON 1
From the MRC Laboratory of Molecular Biology,
Cambridge
SUMMARY
Previous work has shown that multiple HeLa nuclei injected into Xenopus oocytes remain
transcriptionally active for many days and that the expression of HeLa genes in oocytes can
be detected by 2-D gel electrophoresis.
We show here that of 25 proteins which have the electrophoretic properties of HeLa gene
products, only 3 are expressed in injected oocytes. To test that these proteins are products
of HeLa genes, and not products of activated oocyte genes, we have injected HeLa nuclei into
enucleated oocytes. Three days later, several HeLa proteins were synthesized.
The turning off of most HeLa genes in injected oocytes is apparently not at the translational
level. This is indicated by the fact that adenovirus mRNA is efficiently translated when injected into Xenopus oocytes. When adenovirus-infected HeLa cell nuclei are injected into
oocytes the adenovirus genes are not expressed, although some HeLa genes are expressed by
the same nuclei.
The same HeLa genes as are expressed or switched off in injected Xenopus oocytes are also
preferentially expressed or switched off in injected oocytes of a Urodele amphibian, Pleurodeles. This suggests that conditions or molecules may exist in oocytes which selectively impose
on injected nuclei a new programme of gene expression.
INTRODUCTION
Amphibian oocytes injected with nuclei have a potential value for the analysis
of gene control. This has become evident from the recent findings that gene
expression by injected somatic nuclei can be recognized by the direct analysis of
newly synthesized proteins (Gurdon, De Robertis & Partington, 1976 a) or by
enzyme activity assays (Etkin, 1976). The behaviour of somatic nuclei introduced
into amphibian oocytes can be summarized as follows. During the first few days
after injection the nuclei enlarge considerably and exchange their proteins with
those of the oocyte cytoplasm (Gurdon, Partington & De Robertis, 19766).
The rate of RNA synthesis increases as the somatic nuclei enlarge. The message
activity of the RNAs synthesized by the somatic nuclei within the oocytes can
be detected coupled to translation. We reported previously that at least one new
1
Authors' address: MRC Laboratory of Molecular Biology, Hills Road, Cambridge CB2
2QH, U.K.
200
E. M. DE ROBERTIS AND OTHERS
protein could be detected a few days after the injection of HeLa nuclei. The
appearance of this new protein is not due to message carried over with the
nuclei, as shown by injecting RNA isolated from our nuclear preparations. The
appearance of the new protein is eliminated by the injection of a-amanitin
under conditions in which transcription by type B RNA polymerases alone is
inhibited (Gurdon et al 1976 a).
In the experiments reported here we have investigated the selectivity of gene
expression by nuclei injected into oocytes. We wish to know eventually whether
oocytes contain molecules which regulate gene activity. We show here that most
of the identifiable genes expressed in HeLa cells are turned off when HeLa
nuclei are injected into Xenopus oocytes, and that the same spectrum of genes is
inactive in HeLa nuclei injected into oocytes of an unrelated amphibian species.
The injection of purified mRNA indicates that the switching off of gene activity
does not occur at the translational level. We also provide further evidence,
complementary to our previous results, showing that the new proteins observed
are coded for by the injected nuclei and not by the oocyte's nucleus. Enucleated
oocytes were used for this purpose.
MATERIALS AND METHODS
(a) Chemical materials
Most of the chemicals used were 'Analar' grade from B.D.H. Chemicals Ltd.
Urea ('ultra pure grade') was from Schwarz Mann. Proteins used for molecular
weight calibrations were from Sigma. The radioactive isotopes were supplied
by the Radiochemical Centre, Amersham, Proteinase Kby Merck, Deoxyribonuclease I by Worthington Biochemical Co., tissue culture media by GibcoBiocult, and Ampholines by L.K.B.
(b) Biological materials
Oocytes of Xenopus laevis and Pleurodeles waltlii were injected and cultured
as previously described (Gurdon, 1976). After culture at 25 °C for varying
amounts of time, oocytes were labelled and frozen. Samples for protein analysis
were labelled for 6 h in 2 /d/oocyte of medium containing 250 /tCi/ml of L[U-14C]amino acid mixture (57 mCi/mAtom). In every experiment a few
oocytes were fixed and sectioned serially for cytological examination.
HeLa cells were grown as monolayers in Ham's F12 medium supplemented
with 10 % foetal calf serum and as suspension cultures in Eagle's minimal
essential medium with spinner salts and 5 % foetal calf serum. Cultures were
checked periodically for PPLO contamination.
Nuclei can be prepared for injection by several methods (Gurdon, 1976), but
in all the experiments reported in this paper the lysolecithin-bovine serum
albumin (LL-BSA) method of Gurdon (1976) was used. Successful results can
also be obtained with cells ruptured by sucking them into a mjcropipette of
smaller diameter (the method used to transplant nuclei into eggs). Nuclei
Somatic nuclei in frog oocytes
201
prepared with detergents such as Triton X-100 or Nonidet P-40 become pycnotic
after injection into oocytes. About 200 HeLa nuclei were injected into each
oocyte.
(c) Extraction of the samples for protein analysis
Groups of five oocytes were homogenized in 50 fi\ of 10 mM Tris-HCl, pH 7-4,
50 mM-NaCl, 5 mM-MgCl2, 25 /tg/ml deoxyribonuclease I and25 /tg/ml ribonuclease A. After standing for 10 min on ice the yolk platelets were removed
by centrifugation at 1000 g for 5 min. The supernatant was removed, and
the upper lipid layer discarded. After lyophilizing, the supernatant was resuspended in 5/d/per oocyte of 9-5 M urea, 2 % w/v Nonidet P-40, 2%
ampholines pH range 5-7, 5 % 2-/?-mercapto-ethanol, filtered through siliconized glass wool and stored at - 2 0 °C. The presence of salt in the homogenization medium is essential, otherwise several proteins bind to the yolk platelets.
(d) Two-dimensional gel electrophoresis {isoelectric focusing ISDS electrophoresis
The oocyte proteins were separated in the two-dimensional electrophoresis
(2-D) system described by O'Farrell (1975) modified as follows. The first dimension, isoelectric focusing in the presence of urea and Nonidet P-40, resolves
proteins in a pH range from 5 to 7. The 4 % polyacrylamide gels, 12 cm long,
were polymerized inside 1 ml pipettes and the samples run at 400 V for 18 h
in a standard disc-gel electrophoresis apparatus. 40000 cpm of the sample were
loaded on each gel. In our labelling conditions, this usually corresponds to
about 0-5 oocyte.
After focusing, each disc gel was equilibrated in 6 ml of SDS sample buffer
(0-06 M Tris-HCl pH 6-8, 2 % SDS, 5 % A mercaptoethanol, 10 % glycerol)
for 20 min and stored frozen.
The second dimension was run in 15 % polyacrylamide slab gels with the discontinuous buffer system described by Laemmli (1970), using an acrylamide to
bisacrylamide ratio of 200:1 (Knowland, 1974).
After electrophoresis, the gels were fixed in 15 % trichloroacetic acid 40%
methanol for at least 30 min and submitted to fluorography (Bonner & Laskey,
1974). The dried gels were exposed for 1 week on Kodak RP/Royal X-O-mat or XOmatic H film, prefogged to 0-2 O.D. units as described by Laskey & Mills (1975).
In order to refer to the different proteins in the two-dimensional map, each
spot can be located by the use of two parameters. The first is its apparent
molecular weight x 10~3, and the second its apparent iso-electric point expressed
in pH units. For example, the major protein synthesized by HeLa nuclei in
oocytes has a molecular weight of 66000 and an isoelectric point of 6-04 and
is therefore named 66/6-04. The molecular weight calibrations were performed
using /?-galactosidase, bovine serum albumin, catalase, aldolase, chymotrypsinogen and encephalomyocarditis virus-coded proteins as standards, in the same
202
E. M. DE ROBERTIS AND OTHERS
slab gel in which oocyte proteins were electrophoresed. The pH determinations
were performed on parallel isolelectric focusing gels which were sectioned and
equilibrated for 1 h in water before pH measurement, and are less accurate than
the molecular weight determinations (O'Farrell, 1975).
(e) Extraction of cytoplasmic RNA from adenovirus 5-infected cells
HeLa cells were infected with adenovirus 5 at a multiplicity of 100 p.f.u./cell.
The cells were harvested and fractionated into nuclei and cytoplasm 13 h after
infection by the method of Kumar & Lindberg (1972). The cytoplasmic fraction
was then digested for 2 h at 37 °C with 500 /*g/ml of Proteinase K in 50 mM
Tris-HCl pH 7-5, 10 mxi EDTA, 0-5 % SDS (modified from Weigers & Hiltz,
1972, and Gross -Bellard, Oudet & Chambon 1973). The sample was then diluted
into buffer containing 10 mM Tris-HCl pH 7-8, 1 mM EDTA, 0-5 % SDS and
extracted with 1 vol. of 1:1 phenol chloroform (Perry, La Torre, Kelley &
Greenberg, 1972). The aqueous phase was re-extracted with phenol:chloroform
and the first phenol:chloroform phase with 0-1 M Tris-HCl pH 9-0, 0-5 % SDS
(Brawerman, Mendicki & Lee, 1972). The aqueous phases were pooled and
precipitated overnight with 2-5 vols. of ethanol at — 20 °C, after addition of
NaCl to a final concentration of 0-4 M.
RESULTS
(A) HeLa gene expression in frog oocytes
In order to detect the synthesis of HeLa proteins after injection of nuclei into
oocytes, we have used two-dimensional electrophoresis (2-D electrophoresis).
By this method of analysis uninjected Xenopus oocytes (Fig. 1 A) and HeLa
cells (Fig. 1B) show very different patterns of radioactive proteins. This is shown
more clearly in Fig. 1C, in which HeLa and Xenopus proteins were mixed and
coelectrophoresed. At least 25 major HeLa proteins are easily distinguishable
from those of Xenopus oocytes (arrowed in Fig. 1C). Five major proteins coincide in both species (indicated with arrows in Fig. 1B). One of them has been
identified as actin by peptide analysis (Longthorne & De Robertis, unpublished
results) and is the major labelled protein in both types of cells (indicated with an
arrow in Fig. 1 A).
We will now consider the proteins synthesized by oocytes injected with
somatic nuclei. When oocytes are injected with HeLa nuclei and labelled with
[14C]amino acids for 6 h during the first day after injection (Fig. 2 A, B), no new
spots are apparent by comparison with uninjected oocytes (Fig. 1 A). If oocytes
injected with HeLa nuclei are cultured for 3 days and then labelled in [14C]amino
acids for 6 h, at least five new labelled proteins are seen. Of these the three indicated with arrows in Fig. 2D are in the same position as HeLa proteins (in gel
positions 66/6-04, 68/5-92 and 66/5-97 using the nomenclature of apparent
molecular weight x 10~3/isoelectric point). In the case of the two other new
Somatic nuclei in frog oocytes
203
Isoelectric focusing (pH units)
6
7
I
Xenopus oocyte
80-
60-
40-
[HeLa cells
80-|
60o
1
00
40-Wf"^
_
Q
Xenopus+HeLa
mixture
r
-/*•
40-
Fig. 1. Fluorographs of 2-D gels of proteins synthesized by uninjected Xenopus laevis
oocytes and by intact HeLa cells. Proteins were labelled with [14C]amino acids. (A)
Oocyte proteins; the arrow indicates the position of actin. (B) HeLa cells, the arrows
indicate proteins that superimpose with those of Xenopus. (C) Coelectrophoresis of
labelled proteins from Xenopus oocytes (40000 cpm) and HeLa cells (40000 cpm).
The arrows indicate the position of HeLa proteins that do not superimpose with
Xenopus proteins.
204
E. M. DE ROBERTIS AND OTHERS
Isoelectric focusing (pH units)
60-
6-4
6-3
60
i
i
i
a
'
A
%c ,
5-8
i
i
*.
B
*
-70
7
oX
T
U
•S
i>
DayO
Day 0
ao
^((/?
C
60
"
D
a,
• <-•?•** kqpm c
R
JBMZJ
p '
CO
V
W
X
Q
-70
* T
50-
•
Day 3
• Day3
Fig. 2. Fluorographs of proteins synthesized by Xenopus oocytes injected with
HeLa nuclei. Two selected regions of the 2-D gels are shown enlarged. (A) and (B),
Oocytes labelled 2-8 h after injection; (C) and (D), labelled from 72-78 h. The new
proteins induced by HeLa nuclei are indicated with arrows in (C) and (D). Some of
the oocyte proteins are lettered from O to X to help comparison. Oocyte spots which
show variation of intensity from one gel to another are indicated by a and spots near
the edges of the enlarged areas as /?. Oocytes have a faint radioactive background in
region 68/5-9 (in general less than is seen in (B), see Figure 1 A) in the same position
as one of the proteins made by intact HeLa cells. Nevertheless, a significant increase
in radioactivity in this region a few days after injection of HeLa nuclei was observed
in seven independent experiments.
spots (54/6-28 and 55/6-33, Fig. 2C), we were unable to find any equivalent
proteins in extracts of intact HeLa cells, and they could have been coded either
by the HeLa or oocyte genomes.
Protein 66/6-04 is the most abundant of the new proteins, and its appearance
is very reproducible; it has been observed in all nine independent experiments performed using oocytes from seven different frogs. Some of the minor
new spots were not present in all of the 2-D analyses, but this could be due
to insufficient exposure of the gels. Nevertheless, each of the five new proteins
was found in four or more independent experiments. There are several reasons
why we believe that the appearance of these proteins depends on the transcription of new RNA from the nuclei previously injected into oocytes (Gurdon et
al 1976 a). Since the appearance of these spots is sensitive to low concentrations
of a-amanitin (Gurdon et al. 1976a), the RNA is probably synthesized by a
Somatic nuclei in frog oocytes
205
Type B RNA polymerase (Chambon, 1975). HeLa proteins are detected only
when the oocytes are labelled a few days after nuclear injection (and not during
the initial day); this is presumably because in the course of this period the new
mRNA which is synthesized by the transplanted nuclei accumulates in the
oocytes (De Robertis et al 1977).
We can now examine the question of whether all the genes that are expressed
in intact HeLa cells are also active in HeLa nuclei injected into Xenopus
oocytes. As shown in Fig. 1C, at least 25 major HeLa proteins can be readily
distinguished from Xenopus proteins by 2-D gels. Of these only three were
detected in injected Xenopus oocytes, although all 25 would have been detected
if synthesized at the same rate as protein 66/6-04. It therefore appears that the
expression of HeLa genes in oocytes is selective; only a restricted group of
proteins is expressed, while many genes active in HeLa cells become inactive
(or are expressed at rates below the sensitivity of our methods of detection).
(B) Somatic nuclei in enucleated oocytes
In this section we will examine the question of whether the new proteins induced
by HeLa nuclei in Xenopus oocytes are indeed coded for by the injected nuclei
or not. We have previously reported that the major protein induced by HeLa
nuclei in oocytes has CNBr peptides of similar isoelectric points to those of
protein 66/6-04 synthesized by intact HeLa cells (Gurdon et al. 1976a). Similar
results (not shown) were obtained when the peptides were analysed in SDS gels.
Although the CNBr peptides indicate that there are great sequence similarities
between the polypeptide 66/6-04 extracted from injected oocytes and from HeLa
cells, it is nevertheless conceivable that this protein is not of HeLa origin. For
example, when we compared by similar techniques the peptides from the actin
spot of uninjected oocytes with the actin peptides from HeLa cells, they were
found to be similar, a finding that is not surprising since it is known that the
sequence of actin is highly conserved through evolution (Fine & Bray, 1971).
Therefore, it could be argued that the protein 66/6-04 is not coded by the HeLa
nuclei and could be the product of a non-expressed oocyte gene whose transcription is induced by some factor present in the injected material.
We have been able to eliminate this argument using Xenopus oocytes enucleated by the method of Ford & Gurdon (1977). After the wound left by the
manual enucleation had closed, the oocytes were injected with HeLa nuclei,
cultured for 3 days, labelled for 6 h with [14C]amino acids and analysed by 2-D
electrophoresis. The different steps involved in this experiment are shown in
Fig. 3. It was found that protein 66/6-04 is also synthesized in the absence of the
oocyte nucleus. We conclude that protein 66/6-04 does not result from the activation of some previously unexpressed gene of the host nucleus, and that we
are detecting true activity of HeLa genes.
14
EMB 40
206
E. M. DE ROBERTIS AND OTHERS
Oocytes removed
from ovary
"1
200 HeLa
nuclei injected
©
Follicle cell
layers removed
Oocyte nucleus
removed
200 HeLa
nuclei
injected
Oocytes incubated for 3 days for nuclei to swell
Oocytes incubated in [I4C]AA for 6h to label proteins.
Proteins extracted and analysed by 2-D gels
Fig. 3. Schematic representation of the steps involved in the enucleation experiment
described in the text. The follicle cells that surround the oocytes are removed
manually before enucleation, in order to allow the wound left by the removal of the
oocyte nucleus to close (Ford & Gurdon, 1977). Oocytes injected with HeLa nuclei
synthesize spot 66/6-04 (arrowed), although in less amount that non-enucleated
oocytes.
(C) The translation of message for 'unexpressed genes'
As shown in section A, many of the genes active in intact HeLa cells become
inactive (or much less active) in nuclei-injected oocytes. This ' turning off' of
some genes by the oocyte cytoplasm could occur at a transcriptional,
Somatic nuclei in frog oocytes
207
post-transcriptional (processing of the newly made RNA into active message) or
translational level. It seems unlikely that the oocytes would display translational
selectivity since they can translate efficiently many different types of injected
mRNAs (Gurdon, Lane, Woodland & Marbaix, 1971; Lane & Knowland,
1975; Vassart <?* a/. 1975; Lanclos & Hamilton, 1975; Yip, Hew, &Hsu 1975).
However, this question of whether thereis selective translation of certain mRNAs
in oocytes injected with somatic nuclei can be tested directly by injecting mRNA
coding for unexpressed genes.
We have performed this experiment using adenovirus-infected HeLa nuclei,
since it is possible to isolate adenovirus late mRNA in sufficient amounts for
protein synthesis studies. The genes coding for the adenovirus late proteins are
an example of genes that are not expressed in oocytes. In the experiment shown
in Fig. 4 HeLa cells were infected at high multiplicity with adenovirus Type 5
(100 p.f.u. cell) and the nuclei prepared and injected 6 or 12 h after infection.
In the latter situation, the majority of the mRNA that is synthesized in infected
HeLa cells has been shown to be adenovirus late mRNA (Lindberg, Persson &
Philipson, 1972). Most of the proteins synthesized in HeLa cells during late
infection are virion proteins and the 'infected cell-specific polypeptides' (Russel
& Skehel, 1972). All these proteins are very well resolved in 2-D gels (Fig. 4B)
and are readily distinguished from Xenopus endogenous proteins. Nevertheless,
none of these seven adenovirus-induced proteins was detected when oocytes
injected with 6- or 12-h infected HeLa nuclei were examined (Fig. 4A). This
was not due to non-specific damage of the nuclei as indicated by cytological
examination and by the fact that four of the HeLa-induced proteins were detected
in the same experiment (arrowed in Fig. 4A).
In order to test whether the non-expression of the adenovirus genes in oocytes
was due to translational or pretranslational control, we injected adenovirus
mRNA. Cytoplasmic RNA was isolated from HeLa cells (see Methods)
infected 12 h previously with adenovirus, and injected into oocytes. As shown in
Fig. 5, adenovirus mRNA is translated by oocytes. We were able to detect
synthesis of two virion proteins (hexon and fibre in gel positions 110/6-10 and
60/6-4) and of 'infected cell specific polypeptides' 1 and 2 (in gel positions
93/6-08 and 79/5-96). The proteins synthesized in oocytes comigrated in 2-D gels
with the late adenovirus proteins synthesized by infected HeLa cells (Fig. 5,
E and F).
We conclude from these experiments that the observed non-expression of
some genes of somatic nuclei in oocytes is not due to an inability of the oocytes
to translate certain messages (at least in the case of the adenovirus late genes).
Therefore the turning off represents a decreased availability of active mRNAs
due to a decreased rate of synthesis or processing of the gene transcripts.
14-2
208
.
E. M. DE ROBERTIS AND OTHERS
Isoelectric focusing (pH units)
6
7
fl
.
.
........
I
[Oocytes with adeno-infected nuclei
80-
60
r
40o
X
I Adeno-infected HeLa cells
I
60-f
40-
i
•
J
*
Fig. 4. Adenovirus genes are not expressed in oocytes. (A), Proteins synthesized in
Xenopus oocytes injected 3 days previously with HeLa nuclei which had been isolated
12 h after infection at high multiplicity with adenovirus type 5. The four arrows
indicate new proteins, which are the same as those induced by uninfected HeLa
nuclei (see Fig. 2). No adenovirus proteins are detectable. (B), Proteins synthesized by
intact HeLa cells infected with adenovirus. The cultures were labelled with [14C]amino
acids between 12 and 16 h after infection at high multiplicity. Proteins induced by
viral infection are lettered from a to j (a similar gel of uninfected HeLa cells is shown
in Fig. 1B). The viral proteins and 'infected cell specific polypeptides' of adenovirus
type 5 have been studied by Russell & Skehel (1972) using one-dimensional SDS
electrophoresis. Their apparent molecular weight determinations are in very good
agreement with ours. This enables us to tentatively identify the adeno-induced spots
in2-D gels in the following way:
Somatic nuclei in frog oocytes
209
(Fig. 4. continued)
Mol. wt. x 10- 3
Russel & Skehel
(1972)
115
95
79
70
64
62
50
—
Protein
Hexon
ICSP-1
ICSP-2
Penton base
ICSP-3
Fibre
ICSP-4
—
pH
6-4
1
Protein
Fig. 4(B)
a
b
c
d
e
f
g,h
U
Isoelectric focusing
61
1
2-D gel position
110/610
93/6-08
79/5-96
69/700
63/6-3
60/6-4
48/6-23; 48/6-17
36/6-24; 36/6-17
5-8
i
-
100
-
80
60 -
Mock-injected
- 100
x
-
80
+ Adeno mRNA
- 100
60 -
-|-Adeno mRNA
+ Adeno proteins
Fig. 5. Translation of adenovirus mRNAs in Xenopis laevis oocytes. Enlarged
areas of 2-D gel fluorographs. The arrows indicate the relative positions of adenovirus
proteins (hexon, ICSP-1, ICSP-2 and fibre, in order of decreasing molecular
weight). (A, B) Mock-injected oocytes. (C, D) Oocytes injected with cytoplasmic
RNA which had been extracted from HeLa cells infected with adenovirus 12 h
previously (see Methods). (E, F) Coelectrophoresis of oocytes injected with adenovirus
mRNA (40000 cpm of the same sample used in (C) and (D)) together with labelled
adenovirus proteins (10000 c.p.m. of the late adeno-infected HeLa cell sample
used in Fig. 4B). The adeno proteins translated in oocytes have the same 2-D
mobility as in infected cells. For illustration purposes, the fiuorograms were exposed
tofilmfor different amounts of time. This explains why the oocyte endogenous spots
are less intense in some samples (e.g. compare samples (B) and (F)).
80
210
E. M. DE ROBERTIS AND OTHERS
(D) HeLa gene expression in newt (Pleurodeles waltlii) oocytes
As shown in section A, most of the HeLa genes recognizable by 2-D gels
are not expressed in Xenopus oocytes containing HeLa nuclei. An important
question is whether this is a truly selective phenomenon or not. For example, the
active HeLa genes could be those recognized as similar to the genes normally
expressed by Xenopus oocytes, and the 'turned off genes' could be recognized
as those of a kind not expressed by oocytes. Alternatively, the inhibition of most
HeLa genes might have no biological significance, being just a consequence of
the abnormal human-frog cell combination. Although it is difficult to provide a
definitive answer to this question (see Discussion), the probable biological significance (hence selectivity) of the expression or non-expression of HeLa genes
would be greatly increased if we found that the same HeLa genes are turned on
and off in oocytes of another species unrelated to Xenopus.
We used oocytes of a newt, Pleurodeles waltlii, which have a 2-D distribution
of proteins different from HeLa cells. Pleurodeles oocytes are also very different
from Xenopus oocytes when compared by 2-D gels; this is not surprising since
Pleurodeles (being an urodele) is considered to be only distantly related to frogs
(see Nieuwkoop & Sutasurya, 1976). We therefore injected HeLa nuclei into Pleurodeles oocytes. At least three new proteins are detected when the injected oocytes
are cultured for 3 days and then labelled with radioactive aminoacids. These
proteins are not detected when oocytes are labelled immediately after injection
(Fig. 6 A). The most conspicuous of these new proteins coelectrophoresed with
the HeLa spot 66/6-04 (Fig 6B) which is also the major protein induced in
Xenopus oocytes. An increased amount of radioactivity was also found in
region 68/5-92 (arrowed in Fig. 6B), which is in the position of another of the
HeLa proteins expressed in Xenopus oocytes, but this protein is only partially
resolved from an endogenous Pleurodeles spot (the spot labelled ' N ' in Fig. 6).
The second new spot corresponds to one of the major HeLa proteins that is superimposed with a Xenopus protein (position 80/5-52, see Fig. 1B). Unfortunately,
three of the other major HeLa proteins that coelectrophorese with Xenopus
proteins (actin and proteins 55/5-40 and 71/5-48) also coelectrophorese with
Pleurodeles proteins. The third new spot was in position 67/6-13 (Fig. 6B) and
we were unable to find any coincidence between it and a HeLa protein. We
did not detect the synthesis of about 20 major HeLa proteins which are different
from Pleurodeles proteins in 2-D gels, although they would have been detected
Fig. 6. Proteins synthesized in Pleurodeles waltlii oocytes. Fluorographs of 2-D gels.
Some of the Pleurodeles proteins are lettered from K to N in order to help comparison. (A) Ooctyes labelled from 0 to 6 h after injection with HeLa nuclei. (B)
Oocytes labelled 72-78 h after injection with HeLa nuclei; arrows indicate new
proteins. (C) Coelectrophoresis of proteins from uninjected Pleurodeles oocytes with
proteins from labelled HeLa cells; the arrows indicate HeLa proteins.
Somatic nuclei in frog oocytes
0-6 h -
,
m
M
N
K.
72-78 h
M
N
Mixture
-
66/604
L
" ** ''"
^"
211
*"""" ^"
212
E. M. DE ROBERTIS AND OTHERS
if synthesized at a rate similar to protein 66/6-04. Most of these non-expressed
genes are the same ones that are not expressed in Xenopus oocytes.
We conclude from this experiment that the mechanisms that determine the
expression of some HeLa genes and the turning off of others are similar in
Pleurodeles and Xenopus oocytes. In both species the same HeLa gene (66/6-04)
is expressed preferentially and the same HeLa genes are turned off.
DISCUSSION
Gene expression by HeLa nuclei in oocytes appears to be restricted to a
selected group of genes. A few new proteins can be detected by 2-D gels a
few days after injection. These new proteins are dependent upon new transcription within the oocytes (Gurdon et ah 1976a), and are indeed coded for by
the injected nuclei (and not by the host nucleus) as shown by enucleation experiments and by peptide analysis of protein 66/6-04. If all the HeLa genes were
transcribed and translated in oocytes at the same rate as protein 66/6-04 at
least 25 HeLa spots should have been detected, yet only three were detected. It
is possible, however, that other proteins are induced by the nuclei (2-D gels do
not resolve proteins with an isolelectric point above 7), or that some of them are
synthesized at rates below the sensitivity of our methods. In any case, it is clear
that some polypeptides are synthesized in much more abundance than others.
Most of the HeLa genes therefore behave as if they were turned off in Xenopus
oocytes. The cytoplasmic signals which mediate this inactivation are not species
specific since a similar pattern of HeLa gene expression is also observed in newt
{Pleurodeles) oocytes, which are only distantly related to frogs.
The level at which the expression of some genes is turned off has been explored
further using adenovirus-infected HeLa nuclei. After injection into oocytes the
expression of the adenovirus genome is undetectable even though some HeLa
genes are expressed by the same nuclei. By injecting adenovirus late mRNA we
showed that Xenopus oocytes can synthesize efficiently adenovirus proteins (and
also that once synthesized these proteins are not degraded by the oocyte). We
can therefore conclude that the turning off occurs at a step previous to protein
synthesis, at the level of the synthesis or processing of the genetic message.
When nuclei are introduced experimentally into a foreign cytoplasm they
tend to assume the nuclear synthetic activity and nuclear morphology characteristic of the host cell (for reviews see Gurdon & Woodland 1968; Gurdon 1974 and
Harris, 1974). For example, when chicken red blood cells are fused to HeLa
cells, the erythrocyte nuclei reassume RNA synthesis and their chromatin
disperses (see Harris, 1974), or when frog brain nuclei (which normally do not
divide) are injected into Xenopus eggs1 they become active DNA replication
1
A frog egg differs from an oocyte in that it has undergone a hormone-induced maturation process. The germinal vesicle breaks down and the oocyte is released from the ovary.
In contrast to oocytes, eggs are very active in DNA synthesis (after sperm entry or experimental activation) and largely inactive in RNA synthesis (for a review, see Smith & Ecker,
1970).
Somatic nuclei in frog oocytes
213
(Graham, Arms & Gurdon, 1966). Since the cytoplasm is able to modify
nuclear activity, it seems possible that the oocyte cytoplasm could reprogram
gene expression of specific protein-coding genes in the injected somatic nuclei.
Oocyte regulatory molecules could enter the injected nuclei, stimulating^the
transcription of those genes which would be recognized as similar to the oocyteactive genes (or turning off other genes). Although this hypothesis could explain
the selectivity of gene expression displayed by the HeLa nuclei in oocytes, the
experiments described in the previous section cannot establish whether the
oocyte cytoplasm reprograms nuclear expression or not. This is because the
three HeLa proteins detected in oocytes are also synthesized in intact cells (our
analysis would not distinguish whether these genes are being expressed at a similar rate as in intact cells or at a much higher one), and it could be argued that
the selective turning off of genes might not be biologically meaningful (being for
example a consequence of the heterologous nucleus-cytoplasm combinations or
of the experimental manipulations). This question of whether gene expression
by the somatic nuclei is reprogrammed by the oocyte cytoplasm can be answered
by demonstrating the experimental turning on of oocyte-active genes previously
inactive in the somatic cells. This has been achieved recently (De Robertis &
Gurdon, 1977) in experiments which involved the injection of Xenopus somatic
nuclei into newt (Pleurodeles) oocytes, and which show that at least some genes
are indeed reprogrammed.
We thank Dr W. C. Russel for the gift of adenovirus; Drs R. A. Laskey, A. Wyllie and
J. Gottesfeld for advice, and J. Lang and J. Price for technical assistance. E. De Robertis is a
fellow of the Jane Coffin Childs Memorial Fund for Medical Research.
REFERENCES
W. M. & LASKEY, R. A. (1974). A film detection method for tritium-labelled
proteins and nucleic acids in polyacrylamide gels. Eur. J. Biochem. 46, 83-88.
BRAWERMAN, G., MENDECKI, J. & LEE, S. Y. (1972). A procedure for the isolation of mammalian messenger ribonucleic acid. Biochemistry, N.Y. 11, 637-641.
CHAMBON, P. (1975). Eukaryotic nuclear RNA polymerases. A. Rev. Biochem. 44, 613-638.
DE ROBERTIS, E. M. & GURDON, J. B. (1977). Gene activation in somatic nuclei after injection into amphibian oocytes Proc. natn. Acad. Sci. U.S.A., (In the Press.)
BONNER,
DE ROBERTIS, E. M., GURDON, J. B., PARTINGTON, G. A., MERTZ, J. E. & LASKEY, R. A.
(1977). Injected amphibian oocytes: a living test tube for the study of eukaryotic gene
transcription? In Biochemistry of the Cell Nucleus (ed. P. B. Garland & A. P. Mathias).
Biochem. Soc. Symp. 42, 181-191.
ETKIN, L. D. (1976). Regulation of lactate dehydrogenase (LDH) and alcohol dehydrogenase
(ADH) synthesis in liver nuclei, following their transfer into oocytes. Devi Biol. 52, 201209.
FINE, R. E. & BRAY, D. (1971). Actin in growing nerve cells. Nature, New Biol. 234, 115-118.
FORD, C. C. & GURDON, J. B. (1977). A method for enucleating oocytes of Xenopus laevis.
J. Embryol. exp. Morph. 37, 203-209.
GRAHAM, C. F., ARMS, K. & GURDON, J. B. (1966). The induction of DNA synthesis by frog
egg cytoplasm. Devi Biol. 14, 349-381.
GROSS-BELLARD, M., OUDET, P. & CHAMBON, P. (1973). Isolation of high molecular weight
DNA from mammalian cells. Eur. J. Biochem. 36, 32-38.
214
E. M. DE ROBERTIS AND OTHERS
J. B. (1974). The Control of Gene Expression in Animal Development. Oxford and
Harvard University Press.
GURDON, J. B. (1976). Injected nuclei in frog oocytes: fate, enlargement and chromatin dispersal. / . Embryol. exp. Morph. 36, 523-540.
GURDON, J. B. & WOODLAND, H. R. (1968). The cytoplasmic control of nuclear activity in
animal development. Biol. Rev. 43, 223-268.
GURDON, J. B., LANE, C. D., WOODLAND, H. R. & MARBAIX, G. (1971). Use of frog eggs and
oocytes for the study of mRNA and its translation in living cells. Nature, London. 233,
177-182.
GURDON, J. B., DE ROBERTIS, E. M. & PARTINGTON, G. (1976a). Injected nuclei in frog
oocytes: a living cell system for the study of transcriptional control. Nature, London. 260,
116-120.
GURDON, J. B., PARTINGTON, G. & DE ROBERTIS, E. M. (19766). Injected nuclei in frog
oocytes: RNA synthesis and protein exchange. /. Embryol. exp. Morph. 36, 541-553.
HARRIS, H. (1974). Nucleus and Cytoplasm. Oxford University Press.
KNOWLAND, J. S. (1974). Protein synthesis directed by the RNA from a plant virus in a
normal animal cell. Proceedings 13th Intern. Congress Genetics, (ed. E. Carlson). Genetics,
78, 383-394.
KUMAR, A. & LINDBERG, U. (1972). Characterization of messenger ribonucleo-protein messenger RNA from KB cells. Proc. natn. Acad. Sci. U.S.A. 69, 681-685.
LAEMMLI, U. K. (1970). Cleavage of structural proteins during the assembly of the head of
bacteriophage T4. Nature, London, 227, 680-685.
LANCLOS, K. D. & HAMILTON, T. (1975). Translation of hormone induced messenger RNA
in amphibian oocytes. I. Induction by estrogen of messenger RNA coding for vitellogenic
protein in the liver of male African clawed toad (Xenopus laevis). Proc. natn. Acad. Sci.
U.S.A. 72, 3934-3938.
LANE, C. D. & KNOWLAND, J. S. (1975). The injection of RNA into living cells: the use of
frog oocytes for the assay of mRNA and the study of the control of gene expression. The
Biochemistry of Animal Development, vol. 4 (ed. R. Weber), pp. 145-181. Academic Press.
3
14
LASKEY, R. A. & MILLS, A. D. (1975). Quantitative film detection of H and C in polyacrylamide gels byfluorography.Eur. J. Biochem. 56, 335-341.
LINDBERG, U., PERSSON, T. & PHILIPSON, L. (1972). Isolation and characterization of adenovirus messenger ribonucleic acid in productive infection. / . Virol. 10, 909-919.
NIEUWKOOP, P. O. & SUTASURYA, L. A. (1976) Embryological evidence for a possible polyphyletic origin of the recent amphibians. /. Embryol. exp. Morph. 35, 159-167.
O'FARRELL, P. H. (1975). High resolution two-dimensional electrophoresis of proteins. J.
biol. Chem. 250, 4007-4021.
GORDON,
PERRY, R. P., LA TORRE, J., KELLEY, D. E. & GREENBERG, J. R. (1972). The lability of poly(A)
sequences during extraction of messenger RNA from polyribosomes. Biochim. biophys.
Acta. 262, 220-226.
RUSSEL, W. C. & SKEHEL, J. J. (1972). The polypeptides of adenovirus-infected cells. /. gen.
Virol. 15, 45-57.
SMITH, L. D. & ECKER, R. A. (1970). Regulatory processes in the maturation and early
cleavage of amphibian eggs. Curr. Top. in Devi Biol. 5, 1-38.
VASSART, G., REJETOFF, S., BROCAS, H., DINSART, C. & DUMONT, J. E., (1975). Translation of
thyroglobulin 33S messenger RNA as a means of determining thyroglobulin quaternary
structure. Proc. natn. Acad. Sci. U.S.A. 72, 3839-3843.
WIEGERS, U. & HILZ, H. (1972). Rapid isolation of undegraded polysomal RNA without
phenol. FEBSLett. 23, 77-81.
YIP, C. C , HEW, C. L. & Hsu, H. (1975). Translation of messenger ribonucleic acid from
isolated pancreatic islets and human insulinomas. Proc. natn. Acad. Sci. U.S.A. 72,
4776-4779.
{Received 4 January 1977, revised 18 January 1977)